To my great disappointment, after having assembled the kit, I found that the 2 "shoulder" and "Elbow" servos are not powerfull enough to lift the arm (it is too heavy). The servos are MG996 servos.
I tried to make the arm shorter & lighter by removing the silver rod in the middle, along with one of the joints attached to it. (making it 5 DOF). However, the servo is still too weak to lift the arm.
Am I wrong about the MG996 servos not being powerful enough to lift 600grams on a 50cm arm? (Obvuiosly the weight is not centered at the end of the arm)
I would like to fix this by ordering two new servos for the load bearing joints. Can you recommend some good ones?
You should always use servos that have a lot more torque (perhaps 100% more) than is actually needed because servos are not designed to operate continuously at the limit of their force - they will overheat and burn out.
It seems to me that that robot arm has a poor mechanical design. There are three servos mounted near the end of the arm. The correct thing would be to mount all the servos on the base using levers or bowden cables to activate the extremities. A second best option would be to mount the three servos at the top of the final arm where their weight would have the least effect on the forces.
Another (perhaps more practical) option would be to replace the three servos on the arm with much smaller (and lighter) servos.
I have allready tried once to create a lighter arm out of wood, but my physical worksmanship is not good enough (arm was too loose and crooked). That's why I bought a premade "set"
I do not have any metal machining skills. As such, it is unreasonable for me to change the servo mount positions.
It is easier for me to buy 2 new (expensive) servos, then to try and machine the aluminium mounts. (I am a S/W guy, and this is a H/W problem )
quote: "I do not have any metal machining skills. As such, it is unreasonable for me to change the servo mount positions.
It is easier for me to buy 2 new (expensive) servos, then to try and machine the aluminium mounts. (I am a S/W guy, and this is a H/W problem )".
That is why you use fiberglass tubing and similar material!
"Am I wrong about the MG996 servos not being powerful enough to lift 600grams on a 50cm arm? (Obvuiosly the weight is not centered at the end of the arm)"
These types of robotic arms are really only capable of lifting their own weight and not much else (maybe pieces of paper and similar). They are educational devices in that they won't be capable of much beyond movement. For this arm to operate you will need a really good power supply capable of supplying a lot of current at the upper end of the servo operating voltage. What is your current power supply for the arm?
My testing of the plastic geared servos show that they last about 7 days under continuous operation. The gears, begin to wear down, the servos don't hold their position, the feedback circuit works the servos harder till, finally, they just quit.
I have one 25Kg servo that has been under continuous power for 15+ months.
I have one project with 3 metal geared servos, 2 hold a XY platform level and the 3rd is used to turn a electronics package so a laser can scan a room to create an image. These servos are torqued at 5uS per correction.
Another project, with metal geared servos, keeps 2 solar panels on a boom pointing at the sun. These 2 servos are torqued at 1uSec per correction.
Typically, 1 degree of servo torque equals to 11.11uSec of torque.
Metal geared servos come in a variety of rated weights. Metal geared servos should have their own well regulated power supply and do not run well from a breadboard.
I use ESP32's and their MCPWM API for servo torque. ESPServo is a good starter library and can give as low as .3 of a degree of torque.
sillycow:
I do not have any metal machining skills. As such, it is unreasonable for me to change the servo mount positions.
It should not need any metal machining capability to mount the three servos on the base and connect them to the things they operate using wire-in-tube push-pull bowden cables. Push-pull cables are widely used in flying model aircraft - usually with a plastic "wire" sliding in a plastic tube.
If the 6 DOF arm doesn't perform well, you might consider using the servos in a 4 DOF design. Below is a link to a frame kit and a kit that has all the parts including the servos and other parts.
zoomkat:
"Am I wrong about the MG996 servos not being powerful enough to lift 600grams on a 50cm arm? (Obvuiosly the weight is not centered at the end of the arm)"
These types of robotic arms are really only capable of lifting their own weight and not much else (maybe pieces of paper and similar). They are educational devices in that they won't be capable of much beyond movement. For this arm to operate you will need a really good power supply capable of supplying a lot of current at the upper end of the servo operating voltage. What is your current power supply for the arm?
I have both a (5v * 3.5 amp ) and a (9v * 5A).
I would be ok with this arm lifting paper, however: It does not lift it's own weight: not even after I removed 10cm of length by removing the aluminum rod, and a servo.
Robin2:
It should not need any metal machining capability to mount the three servos on the base and connect them to the things they operate using wire-in-tube push-pull bowden cables. Push-pull cables are widely used in flying model aircraft - usually with a plastic "wire" sliding in a plastic tube.
...R
Do you have a link to a similar project? I am having a hard time imagining how to achieve a long enough range of motion to power this arm with cables. I mean, the servo's plastic levers are only 4cm long.
zoomkat:
If the 6 DOF arm doesn't perform well, you might consider using the servos in a 4 DOF design. Below is a link to a frame kit and a kit that has all the parts including the servos and other parts.
sillycow:
Do you have a link to a similar project?
I don't, that's why I mentioned model airplanes - I'm sure there are hundreds of websites about them.
I am having a hard time imagining how to achieve a long enough range of motion to power this arm with cables. I mean, the servo's plastic levers are only 4cm long.
The range of motion will be the same - the servo arm pushes or pulls the inner cable at one end and the other end of the inner cable is connected to whatever the servo arm would have been connected to.
Get your multimeter and measure the voltage supplied to the servos when they are under load. That will tell you how well your power supply is working. At 5v, the servos may be only producing maybe 70% of the torque they would have at 6v. Ebay has 6v UBEC voltage regulators that should work with your 9v power supply. Also a 7805 voltage regulator chip with a diode on the ground lead will output 5.7v.
)